• 대한음성학회지:말소리
• /
• 제68권
• /
• pp.33-47
• /
• 2008

In this paper, we propose a hybrid acoustic and pronunciation model adaptation method based on context dependency for Korean-English speech recognition. The proposed method is performed as follows. First, in order to derive pronunciation variant rules, an n-best phoneme sequence is obtained by phone recognition. Second, we decompose each rule into a context independent (CI) or a context dependent (CD) one. To this end, it is assumed that a different phoneme structure between Korean and English makes CI pronunciation variabilities while coarticulation effects are related to CD pronunciation variabilities. Finally, we perform an acoustic model adaptation and a pronunciation model adaptation for CI and CD pronunciation variabilities, respectively. It is shown from the Korean-English speech recognition experiments that the average word error rate (WER) is decreased by 36.0% when compared to the baseline that does not include any adaptation. In addition, the proposed method has a lower average WER than either the acoustic model adaptation or the pronunciation model adaptation.

• 한국정보통신학회논문지
• /
• 제19권5호
• /
• pp.1047-1054
• /
• 2015

본 논문에서는 음성 인식 성능 향상을 위해 미전사된 음성 데이터베이스를 이용한 효과적인 음향 모델 변환 기법을 기술한다. 본 논문에서 기술하는 모델 변환 기법에서는 기존의 적응 기법을 이용하여 환경에 적응된 GMM을 얻는다. HMM의 가우시안 요소와 유사한 요소를 선택하여 선택된 가우시안 요소의 변환 벡터를 구하고 이를 평균 파라미터 변환에 이용한다. GMM 적응 기반의 모델 변환 기법을 기존의 MAP, MLLR 적응 기법과 결합하여 적용한 결과, 자동차 잡음과 음성 Babble 잡음 환경에서 기존의 MAP, MLLR을 단독으로 사용할 경우보다 높은 음성 인식성능을 나타낸다. 온라인 음향 모델 적응 실험에서도 MLLR과 결합할 경우 기존의 MLLR을 단독으로 사용할 때보다 효과적인 모델 적응 성능을 나타낸다. 이와 같은 결과는 본 논문에서 소개한 GMM 적응 기반의 모델 변환 기법을 채용함으로써 미전사된 음성 데이터베이스를 음향 모델 적응 기법에 효과적으로 활용할 수 있음을 입증한다.

• 말소리와 음성과학
• /
• 제7권3호
• /
• pp.131-138
• /
• 2015

In this paper, we propose L1-norm regularization for state vector adaptation of subspace Gaussian mixture model (SGMM). When you design a speaker adaptation system with GMM-HMM acoustic model, MAP is the most typical technique to be considered. However, in MAP adaptation procedure, large number of parameters should be updated simultaneously. We can adopt sparse adaptation such as L1-norm regularization or sparse MAP to cope with that, but the performance of sparse adaptation is not good as MAP adaptation. However, SGMM does not suffer a lot from sparse adaptation as GMM-HMM because each Gaussian mean vector in SGMM is defined as a weighted sum of basis vectors, which is much robust to the fluctuation of parameters. Since there are only a few adaptation techniques appropriate for SGMM, our proposed method could be powerful especially when the number of adaptation data is limited. Experimental results show that error reduction rate of the proposed method is better than the result of MAP adaptation of SGMM, even with small adaptation data.

• 한국음향학회지
• /
• 제32권5호
• /
• pp.438-445
• /
• 2013

본 논문은 일본어 음성인식기 신규 개발을 위해 초기에 부족한 일본어 음성데이터를 보완하는 방법이다. 일본어 발음과 한국어 발음이 유사한 특성을 근거로 한국어 음성 데이터를 이용한 일본어 음향모델 성능개선 방법에 대하여 기술하였다. 이종언어 간 음성 데이터를 섞어서 훈련하는 방법인 Cross-Language Transfer, Cross-Language Adaptation, Data Pooling Approach등 방법을 설명하고, 각 방법들의 시뮬레이션을 통해 현재 보유하고 있는 일본어 음성데이터 양에 적절한 방법을 선정하였다. 기존의 방법들은 훈련용 음성데이터가 크게 부족한 환경에서의 효과는 검증되었으나, 목적 언어의 데이터가 어느 정도 확보된 상태에서는 성능 개선 효과가 미비하였다. 그러나 Data Pooling Approach의 훈련과정 중 Tyied-List를 목적 언어로만으로 구성 하였을 때, ERR(Error Reduction Rate)이 12.8 %로 성능이 향상됨을 확인하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권7호
• /
• pp.2785-2799
• /
• 2020

We propose a novel speaker adaptation method using acoustic model clustering. The similarity of different speakers is defined by the cosine distance between their i-vectors (intermediate vectors), and various efficient clustering algorithms are applied to obtain a number of speaker subsets with different characteristics. The speaker-independent model is then retrained with the training data of the individual speaker subsets grouped by the clustering results, and an unknown speech is recognized by the retrained model of the closest cluster. The proposed method is applied to a large-scale speech recognition system implemented by a hybrid hidden Markov model and deep neural network framework. An experiment was conducted to evaluate the word error rates using Resource Management database. When the proposed speaker adaptation method using i-vector based clustering was applied, the performance, as compared to that of the conventional speaker-independent speech recognition model, was improved relatively by as much as 12.2% for the conventional fully neural network, and by as much as 10.5% for the bidirectional long short-term memory.

• 한국음향학회지
• /
• 제38권1호
• /
• pp.47-50
• /
• 2019

본 논문에서는 잡음 환경에서 효과적인 음성 인식을 위하여 DNN(Deep Neural Network) 기반의 잡음 오염 함수 예측을 이용한 음향 모델 적응 기법을 제안한다. 깨끗한 음성과 잡음 정보를 입력으로 하고 오염된 음성에 대한 특징 벡터를 출력으로 하는 DNN을 학습하여 비선형 관계를 갖는 잡음 오염 함수를 예측한다. 예측된 잡음 오염 함수를 음향모델의 평균 벡터에 적용하여 잡음 환경에 적응된 음향 모델을 생성한다. Aurora 2.0 데이터를 이용한 음성 인식 성능 평가에서 본 논문에서 제안한 모델 적응 기법이 기존의 전처리, 모델 적응 기법에 비해 일치, 불일치 잡음 환경에서 모두 평균적으로 우수한 성능을 나타낸다. 특히 불일치 잡음 환경에서 평균 오류율이 15.87 %의 상대 향상률을 나타낸다.

• 말소리와 음성과학
• /
• 제6권2호
• /
• pp.29-34
• /
• 2014

In the Vector Taylor Series (VTS)-based noisy speech recognition methods, Hidden Markov Models (HMM) are usually trained with clean speech. However, better performance is expected by training the HMM with noisy speech. In a previous study, we could find that Minimum Mean Square Error (MMSE) estimation of the training noisy speech in the log-spectrum domain produce improved recognition results, but since the proposed algorithm was done in the log-spectrum domain, it could not be used for the HMM adaptation. In this paper, we modify the previous algorithm to derive a novel mathematical relation between test and training noisy speech in the cepstrum domain and the mean and covariance of the Multi-condition TRaining (MTR) trained noisy speech HMM are adapted. In the noisy speech recognition experiments on the Aurora 2 database, the proposed method produced 10.6% of relative improvement in Word Error Rates (WERs) over the MTR method while the previous MMSE estimation of the training noisy speech produced 4.3% of relative improvement, which shows the superiority of the proposed method.

• 한국음향학회지
• /
• 제30권6호
• /
• pp.315-323
• /
• 2011

본 논문은 maximum a posteriori linear regression (MAPLR) 기반의 고속 화자적응 성능을 개선하기 위하여 사전분포를 추정하는 두 가지 방식을 제안한다. 일반적으로 MAPLR 방식에서 사용되는 변환행렬의 사전분포는 화자독립모델을 구성하는 훈련 화자들로부터 추정되어 모든 화자들에게 동등하게 적용된다. 본 논문에서는 새로운 화자에게 보다 더 적합한 사전분포를 적용하고자 적응 데이터를 이용하여 새로운 화자의 음향특성과 가까운 참조화자 집단을 선택한 후 참조화자 집단으로부터 사전분포를 추정하는 방법을 제안한다. 또한, 블록 대각 형태의 변환행렬의 사전분포를 추정하는 경우 사전분포의 평균행렬과 공분산행렬을 동일한 훈련 화자들로부터 얻어진 두 가지 형태의 변환행렬집단으로부터 각각 추정하는 방법을 제안한다. 제안된 방법의 성능 평가를 위하여 고립단어 인식실험을 통해 적응 단어의 개수에 따른 단어 인식률을 평가한다. 실험결과, 적응 단어 수가 매우 적을 때 기존의 MAPLR 방식에 비하여 통계적으로 유의미한 성능향상이 얻어짐을 보여준다.

• ETRI Journal
• /
• 제33권3호
• /
• pp.423-433
• /
• 2011

This paper introduces an adaptive and integrated utterance verification (UV) framework using minimum verification error (MVE) training as a new set of solutions suitable for real applications. UV is traditionally considered an add-on procedure to automatic speech recognition (ASR) and thus treated separately from the ASR system model design. This traditional two-stage approach often fails to cope with a wide range of variations, such as a new speaker or a new environment which is not matched with the original speaker population or the original acoustic environment that the ASR system is trained on. In this paper, we propose an integrated solution to enhance the overall UV system performance in such real applications. The integration is accomplished by adapting and merging the target model for UV with the acoustic model for ASR based on the common MVE principle at each iteration in the recognition stage. The proposed iterative procedure for UV model adaptation also involves revision of the data segmentation and the decoded hypotheses. Under this new framework, remarkable enhancement in not only recognition performance, but also verification performance has been obtained.

• 한국자동차공학회논문집
• /
• 제4권4호
• /
• pp.179-189
• /
• 1996

An unstructured grid finite-volume method has been applied to predict the linear and nonlinear attenuation characteristics of the expansion chamber silencer system. In order to achieve a grid flexibility and a solution adaptation for geometrically silencer system. In order to achieve a grid flexibility and a solution adaptation for geometrically complex flow regions associated with the actual silencers, the unstructured mesh algorithm in context with the node-centered finite volume method has been employed. The present numerical model has been validated by comparison with the analytical solutions and the experimental data for the acoustic field of the concentric expansion chamber with and without pulsating flows, as well as the axisymmetric blast flowfield with open end. Effects of the chamber geometry on the nonlinear wave attenuation characteristics is discussed in detail.